Dark Flats / Flats

[Droogie 2001]

Hi,

Sorry for another Flats /Dark Flats thread, I wonder what can of worms I am opening up for myself...

Some background first

I use an illuminated Aurora Flatfield panel to take Flat frames with my setup.
I have also used SGP's "Flat Frame Calibration wizard" to identify the exposure time with an ADU of around 23000 for which is within the suggested range for my CCD.
I find that this does a good job at helping to produce a cleanly calibrated pre-processed stacked image, all good.
Now I have RGB filters and my Luminance channel which is in fact an empty filter wheel spot which uses a IDAS P2 LP filter in the main imaging train. (all of the filters are subjected to the IDAS filter as it is at the beginning of my imaging train)

Onto my question which is about Dark Flats rather than simply Flat Frames.

Am I correct in saying that the Dark Flat frames need to be taken at the same exposure time that the regular Flat frames are taken at with each individual filter?

For example:

To get the required ADU rating on my Luminance channel it takes 7.2 secs exposure time.
To get the required ADU rating on my Red channel takes a longer exposure time 54 secs.
The other Green and Blue filters are 21.97 and 13.87 respectively.

Therefore do Dark Flats need to taken with each individual Filter to compensate for the different exposure times, or is this not relevant? If not what exposure time should I aim for?
I stack with DSS and having read the Help guide it does fully support BIAS and Dark Flats frames, it refers to BIAS, Dark, Dark Flat and Flat as the 'Full Calibration Process'. There are suggestions of using a number of alternate processes which omit either the Dark Flats or BIAS frames.

Having read many threads about Flats, Darks and BIAS I know I am entering a realm of many differing opinions and tastes.
Such as using Master BIAS, Master Darks or not using Darks or BIAS etc. Rather than going down this route I am curious on my current setup. Of course I am always keen to learn different techniques but just trying to keep things simple. After all creating a batch of Dark Flats is not a challenging task with the UK weather…

As I say my Flats seem to work well for me so you may ask why both changing anything. Being new to CCD imaging still means I have a whole lot to learn and so I keen to make sure I am not making some glaring error which will affect the overall quality of my final images.

In particular I have wondered about by my regular Flats exposure times for say the Red filter channel being 54 seconds which seems excessive? I appreciate the red filter receives less light so it makes sense and I would suspect an HA filter could take even longer still?
That’s said my SGP produces the correct ADU (Circa 23000 ADU) for my specific CCD so I don't see it being a problem?

Apologies if I have waffled but I cannot seem to find a straight forward answer to this question.

Many Thanks

[wimvb]

Experiment. As you wrote, considering the weather, getting the dark flats is no extra effort. See it as an opportunity to get to know your gear and conditions.

Sinxe your process works as is, the result may be that it doesn't make much difference. But that's experience gained.

[Jokehoba]

In a word, yes, you are correct: "...Dark Flat frames need to be taken at the same exposure time that the regular Flat frames are taken at with each individual filter."

The term 'dark flat' is bit misleading though. It's really a Flat Dark (i.e., a dark to subtract from a flat frame). But you knew that...

Typically people use Bias as a close approximation to a true 'flat dark' - to save time - since flats are usually of short duration. Yours are quite long exposures. Could you reduce your exposure time by increasing the panel brightness? That would make doing flat darks unnecessary. I guess whether a flat is 'short' is relative the the Light exposure duration - a 20s flat is short if the light frame is 20mins but more significant if your lights are 60s.

But then, whether you use bias/flat darks depends upon how you calibrate your sub-frames - dark subtraction vs BPM.

[ollypenrice]

I share Jokehoba's concern with your flat exposures. My standard answer would just be to say, 'Use a master bias as a master dark for all your flats.' In short exposures your master bias will be insignificantly different from a dedicated dark at the same exposure time as your flats. Most of us do this and taking dedicated darks for flats is a waste of time.

But I hesitate in your case because of the extremely long subs you are needing to make flats. Mine rarely take more than a second. On shutter cameras you have to slow that down or you record the wipe of the shutter. But 54 seconds is bizarre for a red flat. I would check the panel. At these exposure lengths you might be getting thermal noise as well as bias signal in which case a master bias might not work properly as a dark for flats.

Olly

[Davey-T]

Just increase the panel brightness and reduce exposure times to achieve the same ADUs, you don't need to use the same panel brightness for each filter.

Dave

[Droogie 2001]

Thanks for the replies. My panel has no brightness control it just on or off. 

The panel doesn't have any extra filters to reduce the brightness so I am at a bit of a loss as to why it's takes a while. The panel seems okay. I'll run some tests outside of SGP and check the ADU. Maybe SGP is taking excessively long for some reason. Good to hear confirmation on the dark flats filter though this maybe prove irrelevant if the exposure time is the same for all filters should I identify an issue with my exposure times.

Perhaps my IDAS LP filter is the culprit?

Will report back...

[ollypenrice]

1 hour ago, Droogie 2001 said:

Thanks for the replies. My panel has no brightness control it just on or off. 

The panel doesn't have any extra filters to reduce the brightness so I am at a bit of a loss as to why it's takes a while. The panel seems okay. I'll run some tests outside of SGP and check the ADU. Maybe SGP is taking excessively long for some reason. Good to hear confirmation on the dark flats filter though this maybe prove irrelevant if the exposure time is the same for all filters should I identify an issue with my exposure times.

Perhaps my IDAS LP filter is the culprit?

Will report back...

Could it be your power supply? My experience with Aurora panels was dire, with the brittle cables failing at every possible opportunity - wherever they joined either the panel or the power supply. After binning three of the wretched things I said 'Never again' and now use a graphics EL panel which works fine.

It really should not take more than a few seconds to take a flat. Sure, try without the LP filter because I have never used one of these and don't know what they block, but lots of people use them and the nearly one minute flat is certainly a new one on me... This can't be right.

Olly

[Davey-T]

Try using a computer screen set to plain white to see if it's any better.

I've used this method in the past and had to put sheets of printer paper over it to reduce the brightness so if you still get long flats using this method something is amiss somewhere.

Dave

[Droogie 2001]

Thanks again for all of the advice.

Okay so I removed a black neutral density film and tried again in the SGP Calibration wizard.
My Red Filter has dropped to 0.75 seconds which is considerably quicker than the previous 54 seconds.
The ND filter it came with is dark but I am surprised it caused such a difference in time, maybe it was set for some ultra fast system? The issue I have now is that they are too quick as it reaches the camera limit with Luminance.

i.e. It cannot obtain 25000ADU in enough time.
So I need to make the panel duller. Any ideas ideally I need a ND filter with a very slight dullness to them. I'll probably speak to the guys I bought the panel from.
 

[Davey-T]

Making flats is much easier with a dimmable panel, you can experiment with white paper over it.

Dave

[ollypenrice]

9 minutes ago, Davey-T said:

Making flats is much easier with a dimmable panel, you can experiment with white paper over it.

Dave

I have one with a dimmer but find it incosnsitent and some dimmers cause flicker. I find it best to run it flat out and use typing paper. Obviously if it works use it, though.

Droogie, I don't understand what an ND filter was doing in the story? Where did it go?

Olly

[Droogie 2001]

The black sheet which I think is ND film sat sandwiched between the plastic white outer panel and the actual light source . When I enquired about the panel originally I gave them my specifications. The said I may need more filter due to my f4.9 FR.

i am not sure if the sheet was added by mistake or was used by someone previously. Anyway at least I have a means of working out a solution. Like Davey says I can try and add a sheet of paper/s to help guage what amount of dimming I need to aim for. 

Thanks all with helping me solve the problem (or shall I say shedding light on the matter ?)

[Stub Mandrel]

Embarrassing question ... I think I may have got the idea from this thread, but just to be clear what is a dark flat?

Is it a dark of the same exposure as my flats?

 

[Droogie 2001]

Yes that's correct.

[ollypenrice]

4 hours ago, Droogie 2001 said:

Yes that's correct.

Well, it is if you have a lot of time to spare! If you just want a perfectly good dark flat that will be perfectly effective it can simply be a master bias. This will do the job. I don't know anybody who sits through the palaver of shooting 25 flats at 1.002 sec exposure then does a corresponding dark set at 1.002 sec exposure. Honestly this would be a waste of time. A master bias will do the job. I used to do it until the great and the good of this forum told me not to and explained why (thank you Narrowband Paul) and now I don't waste my time shooting darks for flats. I just use a master bias. Try it. It worls.

Olly

[Stub Mandrel]

27 minutes ago, ollypenrice said:

I just use a master bias. Try it. It worls.

Olly

I already have, because I didn't know what a dark flat was, I couldn't use them...

[Droogie 2001]

12 hours ago, ollypenrice said:

Well, it is if you have a lot of time to spare! If you just want a perfectly good dark flat that will be perfectly effective it can simply be a master bias. This will do the job. I don't know anybody who sits through the palaver of shooting 25 flats at 1.002 sec exposure then does a corresponding dark set at 1.002 sec exposure. Honestly this would be a waste of time. A master bias will do the job. I used to do it until the great and the good of this forum told me not to and explained why (thank you Narrowband Paul) and now I don't waste my time shooting darks for flats. I just use a master bias. Try it. It worls.

Olly

I think now I have identified an issue with my flat panel in that it is not sending out enough light I will resort to a master BIAS once I have resolved this. My current red flat takes 54 seconds as opposed to 0.5 seconds for my BIAS so a huge difference. 

At least this thread has identified an issue with my setup and one which should be easy to resolve. Thanks.